Our long-term research aspiration, which we hope will set the GDMC apart from the other research groups in this field around the world, is the work on a '5D super model' (3D + time + scale) for both discrete and continuous spatial phenomena as a basic model concept for SII (Spatial Information Infrastructure, also see SII). This would then provide the basis on which all future models would be founded. To achieve this end, the models will have to become more and more formal (e.g. by increased stress on ontology). The main features of the SII based on a DBMS with the 5D super model at its core, to be realized in the long term, are integration of the various data sets involved (e.g. indoor/outdoor, surface/subsurface, designed/surveyed), real-time or at least 'daily fresh' maps (via sensor web + formal semantics + automatic processing), combination with positioning and wireless communications technologies (to enable the system to deal with moving objects).
Background
The main innovations relate to thinking in terms of infrastructure when developing geoinformation handling systems (which requires technological innovations enabling optimum use of geo-data and services) and increased focus on modelling aspects and the geographical man-machine interaction. The foundation of SII consists of geo- DBMSs filled with geographical data. We have formulated a consistent GIS technology research programme concentrating on the central theme of a geo-DBMS. All research projects will have geo-DBMS roots and from this angle cover such topics as 3D spatio-temporal modelling, computational geometry (spatial data structures/ algorithms), distributed GI processing (network protocols/ interoperability/web services/cloud computing), mobile GIS (LBS) and knowledge engineering (ontology and semantics).
In addition to modelling 2D and 3D spatial aspects of features, the modelling of changes in features over time becomes increasingly important. This will depend on the nature of the changes involved, which may be discrete (e.g. ownership of a parcel) or continuous (e.g. movement of dunes). The integration of this fourth dimension into existing data structures such as rasters, TINs and DEMs, represents a major research topic involving a wide range of issues, from the translation of logical and conceptual models described in the Unified Modelling Language (UML) to the implementation of the associated physical model in a DBMS. The level of detail (or scale) should be treated together with the other spatial and temporal dimensions to yield a 5D super model. In the past, spatial data management has been handled by GIS software outside the DBMS. As DBMSs are nowadays spatially enabled, GISs have migrated (or will soon migrate) towards an integrated architecture: all data (spatial and thematic) is stored and managed in one DBMS. As stated above, the next step will be the creation of a shared SII for related organizations, the so-called information community. It will be a major challenge to integrate the temporal and scale aspects into the traditional 2D/3D models to replace their separate temporal and scale treatment.
The realization of the '5D super model' is a huge challenge. By first extending the traditional 2D models in different dimensions (third spatial dimension, temporal dimension, scale dimension) independently, important experience will be gained with the resulting different types of 3D models based on research which should be well feasible and in reach. Besides developing these advanced models, it will also be investigated how these models can be realized within in an extended spatial DBMS environment. Some of the resulting models are even already operational, e.g. 3D space, though there is still room for improved modeling; e.g. support for true 3D topology is lacking. Next the combination of two extended 3D models will be explored; e.g. 3D space and time, or 3D space and scale, or 2D space and time and scale, resulting in different types of 4D models. These currently belong to hardly explored types of models, but also these are expected to be feasible (in part) and give more insight in complex integrated modeling. Finally, using all these experiences the '5D supermodel' is aiming at integrating 3D space with both time and scale. Currently we do not know whether and how such a model could be developed. The next step after that will be investigations aimed at expanding the DBMS by addition of the capabilities needed to handle this integrated 5D data model.
Research issues and questions
The importance of the geo-DBMS is increasing with the transition to the SII, because not just one organization depends on it, but the geo-information community as a whole. The main use will be query oriented and less update oriented. Only one organization will be responsible for updating a specific type of data, all others query and use these data. As the query aspect is dominant and the cost of memory chips falls, VLM (Very Large Memory) DBMSs might offer a very suitable technical solution. They are powerful enough to serve large numbers of users via the Internet. Further work is needed to increase the bandwidth of the network infrastructure to get the data to the (mobile) users in time. Developments in hardware, software and database technology will contribute to the future shape of the geoinformation infrastructure. Current extensible DBMSs are very capable of storing 2D spatial data. Simple queries like zooming in and zooming out can also be handled efficiently by making use of spatial indexing and clustering. However, more complex operations like map overlay, on-the-fly generalization, enforcing correct topology during updates and 3D analysis are not yet within the reach of these systems. New developments in DBMS technology, like extensible (object relational) DBMSs, object-oriented DBMSs and VLM databases, will underpin the new generation of DBMSs. At the same time, data sets are being made bigger and bigger; this poses serious challenges in DBMS management, such as the handling of point clouds in AHN2 (actual height model of the Netherlands, version 2) or high-resolution imagery.
The SII discussed here is based on interoperability standards and advanced geo- DBMS technology, and can be used to re-implement existing GIS applications in a more efficient and effective manner. It will integrate sectors that have been hitherto separate. Apart from re-using general geoinformation knowledge (perhaps first discovered in one domain and then applied in other domains), the SII will also give an impulse to the use of geoinformation and services from other domains. A prerequisite is that these different domains 'understand' each other, i.e. share a common ontology. It is already difficult enough to share the concepts within a single geoinformation domain (transportation, topography, geology, groundwater, surface water, land registration, elevation, land use, utilities etc.), so one can image the difficult task ahead when concepts have to be shared between different domains. Finally, the SII can be used to build new, sophisticated applications such as mobile location dependent applications, VR (Virtual Reality) and AR (Augmented Reality). As stated above, we are in the middle of a number of fundamental developments. However, there are still many open problems. The GDMC aims to help to solve these problems by carrying out research on the standardization of complex features, the inclusion of quality in the DBMS (perhaps at data type level), 3D modelling and visualization and the usability of VLM databases for GIS applications.
As stated above, the most important innovations in GIS technology research are characterized by conceptual exploration of the infrastructure needed for handling geoinformation. By analysing specific GIS applications, our researchers identify deficiencies in the current technology and knowledge gaps. Scientific research provides a basis for the development of improvements and solutions which are then tested in practice. Developed (prototype) solutions will be made available as open source software. Making GIS technology research results available in the form of freely accessible open source software is, from the academic point of view, perhaps as important as publishing a scientific paper. It allows others to evaluate the new developments and to build further on them. It works in two directions: The GDMC has used open source software made available by other researchers to jump-start the development of our new prototypes and then made the software we developed available as open source - thus allowing others to reuse our ideas while at the same time enhancing our academic reputation. We therefore intend to make all the software we develop available as open source, and to refine this approach in the near future by providing easier access, more documentation and better version management. Research at the GDMC is divided into the two application-oriented theme groups discussed below, which share several types of geo-ICT base technologies.
Theme groups (application-oriented research)
Crisis management: The theme group on Geoinformation for crisis management focuses on building advanced frameworks, developing solutions and testing prototypes that permit knowledge-based use of geoinformation to assist the decision- making process in crisis situations, taking into account the time constraints, communication, network and visualisation limitations. The research of the theme group is derived from, and therefore shared with, the fundamental technological research on geo-DBMSs performed at the GDMC.
Spatial information infrastructure: The SII theme group focuses on the technological aspects of the realization of a spatial information infrastructure. One important prerequisite here is the establishment of key data sets, such as those containing topographic and land registration information. SII developments are currently ongoing at various levels such as the set of authentic registrations (and the relationship with Dutch standard NEN3610 on base model geoinformation) in the Netherlands, INSPIRE at European level and the Land Administration Domain Model (LADM - ISO/TC211) at global level.
Basic Geo-ICT technology
Core research theme:
Geo-DBMS: The GDMC carries out fundamental research on the use of spatial data types, operators, functions, clustering and indexing in Database Management Systems. The geo-DBMS is becoming increasingly important in the transition to the SII, because not only one organization but the entire geo-information community depends on it. The research topics within this theme are topology structure management within the DBMS, the handling of 3D, temporal and dynamic objects within the DBMS, large point cloud data sets, comparative functional and performance benchmarks and XML (eXtensible Markup Language) support at the DBMS level.
Additional research themes, all linked to geo-DBMS:
3D spatio-temporal modelling: This research topic focuses on the challenges related to (static and dynamic) data modelling in various systems (such as geo- DBMS, GIS and CAD) and the investigation of new representation and modelling concepts. Research on 3D spatial modelling and 2D spatiotemporal modelling is ongoing, and some preliminary research on 3D spatiotemporal modelling has been started. Furthermore, various aspects of data integration and data harmonization are undergoing extensive investigation.
Computational geometry (spatial data structures/algorithms): Linking GIS to computational geometry, spatio-temporal modelling and simulation models permits simpler, faster, more powerful and flexible use. An important related theme is generalization. The temporal component plays an important role in these models, leading to the development of dynamic geographic information systems.
Distributed GI processing (network protocols, interoperability and web services): This theme emphasizes research in the field of distributed GISs, data transfer between various systems, interoperability, geoinformation standards, spatial models and query languages. Geoinformation processing will need to be subjected to geodetic quality control in the phases following data capture, such as data modelling, analysis and visualization. Such aspects as components and storage of quality, meta data and error propagation also have to be taken into account.
Mobile GIS (LBS): Mobile GIS or Location Based Services involve the integration of at least three types of technologies: positioning (GPS, Galileo), wireless communication (GSM, GPRS, UMTS) and GIS (geo-DBMS, geocoding, routing, user interface, small-display cartography). Due to the dynamic and mobile aspects, this type of environment brings obvious potential benefits to a number of applications (navigation/travel support, localized news services, traffic and fleet management, field observations and data collection, etc.), but also has its own research challenges (in particular the architecture and design of these systems).
Knowledge engineering (ontology and semantics): Agreeing on the syntax and formats of spatial data and the development of systems to be used for handling such data is the first step towards interoperability. But getting the syntax and format right does not yet mean that we understand one another's information - the essential next step is to reach agreement on the domain (or thematic) models to be used. Study of the semantic aspect of information is not only important to help human beings to understand each other, but is also essential if we want machines to process this information in useful ways. To this end, the semantics will have to be formalized with the aid of semantic webs, ontologies, etc. OWL (Ontology Web Language) is a useful new tool in this field.
Supporting research themes and technologies:
Spatial data capture
Positioning and geo-information
Geo-visualization
Last edits 2015-02-11